There are numerous systems and methods known in the art to determine color of an object, and depending on the particular application or type of object, color determination methods and systems may vary substantially. For example, many medical and other scientific applications prefer hyperspectral imaging, especially where precise measurement is necessary. Hyperspectral imaging typically provides spectral analysis (often with a resolution of less than 10 nm) of each pixel in a field, thereby generating excellent data on the object. For example, based on such imaging, tissue demarcations (such as may be present in neoplastic or other pathological conditions) may be readily identified.
In contrast, all or almost all known cosmetic applications exhibit a significantly less sophisticated degree of analysis. For example, MacFarlane et al. describe in U.S. Pat. No. 6,067,504 an analytic system in which color is averaged over an entire field, rather than on a pixel-by-pixel basis. Consequently, a major drawback of such systems is that the processes of averaging frequently provide false readings where one or more hair is present having different colors. For example, one person having some dark brown hair and some white hair would be found to have the same hair color as another person having all light brown hair. Yet the cosmetic effect of the two would be entirely different. Moreover, the two persons would likely need different hair color products to achieve the same end coloration.
More recent patent applications, WO 01/55956 and U.S. Ser. No. 09/493,511 (filed Nov. 3, 2000 and Jan. 28, 2000, respectively, and both incorporated by reference herein) describe analysis of subsets of a field during cosmetic color analysis of hair or skin. Such systems advantageously permit substantially increased accuracy of ‘real-life’ color using an ordinary video or other digital camera, as opposed to an expensive colorimeter. For example, the subsets can be analyzed down to a pixel-by-pixel level, which typically allows the system to compensate for shading, glare, and other effects.
However, despite improved color analysis, such systems still provide output as a single color reading. Therefore, blotchy skin or highlighted areas in an otherwise homogenous hair population will nevertheless be presented as a single color output. For example, on a descriptive scale the computed color may be presented as “light blonde”, “auburn”, or “silver”. Alternatively, on a numeric scale the color may be presented as 1.78, 4.22, or 7.29.
Thus, although there are various methods and systems known in the art to determine the color of an object, and particularly of a cosmetic object, all or almost all of them suffer from one or more disadvantages. Therefore there is still a need to provide improved methods and systems for cosmetic color determination.